CLARIN Café | TranscriptionPortal – User-Friendly Speech Recognition, Translation and Summary
- Date
- October 29, 2026
- Location
- Online
General Information
Spoken language is the most important carrier of information in many research areas, in field work and in everyday life. The TranscriptionPortal makes this information accessible via a web service. It’s as easy as 1-2-3: 1) upload your audio, 2) select the input and output languages, 3) start processing – and watch how your data makes its way through the matrix.
The TranscriptionPortal features a modern graphical user interface, and it currently supports five languages: English, Dutch, Italian, German and Greek. The typical workflow consists of automatic speech recognition, followed by optional manual validation, translation and summarization. For maximum usability for non-technical users, the individual services are preconfigured, so that only a very small set of options needs to be set. The outcome is a downloadable set of text files with the transcribed and translated text, in full length or as a summary.
The use of the TranscriptionPortal is free for academic institutions. To ensure high standards of privacy, data is stored only during processing, all communication is encrypted, and only trusted academic service providers are accessed.
The TranscriptionPortal is being operated by the Bavarian Archive of Speech Signals in close collaboration with CLST in Nijmegen, and is supported by CLARIN. In November 2025, it was evaluated successfully in a usability test.
- Time: 14:00 - 16:00 CEST
- Venue: online
A full overview of planned and previous CLARIN Café sessions can be found on the CLARIN Café page.
Target Audience
All welcome, though the Café may be particularly relevant to researchers in linguistics and language sciences, digital humanities, LT and NLP, social sciences, speech technology, and practitioners in language teaching.
To help us tailor the Café, please provide details about your own background and your questions upon registration.
Programme
14:00 - 14:15 Introduction
14:15 - 14:30 demo
14:30 - 14:40 Recent new features (results from ATRIUM research Forum
14:40 - 14:50 Use case preferably from participants
14:50 - 15:00 Ideas for new features and improvements
15:00 - 15:10 Summary and comparison with other solutions
Speakers
Christoph Draxler is the head of the Bavarian Archive for Speech Signals at LMU Munich. He received his PhD in Logic Programming and Databases at Zurich University, and since then works as a researcher at the Institute of Phonetics and Speech Processing at LMU Munich. His main research interests are speech databases, transcription, and the development of software for speech technology.
Henk van den Heuvel is a senior researcher and Head of the Humanities Lab at Radboud University. His work centers on language and speech technology, speech transcription, research data management, and the development of language resources for both research and AI applications. He is also a member of CLARIN’s Board of Directors.